Learning Style Prediction Using Students’ E-book Reading Behaviors Data

Abstract

Adaptivity is one of the most prominent features of intelligent textbooks in the 21st century. Learning style is a personality characteristic of learners, which is used to describe learners' preference for processing information in a certain way. Learning style was often measured by questionnaires, which were easily influenced by learners' subjective cognition and external interference. This study proposes a data-driven approach to automatically detect learning style of learners. In the learning environment of e-textbook, 234 students' reading data was collected, and a learner model is constructed using machine learning technology. The results show that the proposed model achieves a promising performance in prediction learning style. This will help measure learning style more accurately and provide support for personalization. The learner model applied to e-textbook can promptly and dynamically monitor the changes of students' learning behavior in the online environment, and adaptively intervene, remedy or enhance.